JP2006319758A - Communication device, communication system, and communication program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a communication device which performs trouble restoration of a post-establishment type of high speed in a low cost. <P>SOLUTION: The device is provided with a path administrator 55 which memorizes present path information regarding to the present path in which an own device becomes a data transferring path for the present path disconnected by a trouble in a communication system, at a node 16 in the communication system performing connection type communication; a relief path route selector 57 which selects and memorizes relief path information, regarding to the relief path corresponding to the present path which is established as a path of data transfer instead of the present path which is memorized in the path administrator 55 before the present path is disconnected when the present path is disconnected; and a trouble restoration administrator 54 which transmits the relief path information corresponding to the present path, which the relief path route selector 57 selects and memorizes to the node on the present path which is disconnected when the present path is disconnected. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a communication device, a communication system, and a communication program that perform communication path failure recovery.

  As a packet exchange method and a method for establishing a connection for packet exchange, a communication method called MPLS (Multi-Protocol Label Switching) is used in communication networks and the like. Also, MPLS-enhanced SONET / SDH (Synchronous Optical NETwork / Synchronuos Digital Hierarchy) network cross-connect devices (switching devices), WDM (Wavelength Division Multiplex) network OXC (Optical Cross Connect) devices, etc. As a method for establishing a connection in circuit switching, a communication method called GMPLS (Generalized MPLS) has been standardized and applied to experimental facilities and some communication networks.

  In MPLS and GMPLS, in order to establish a communication path between two remote points providing a communication connection as an LSP (Label Switched Path) (hereinafter referred to as a path), communication devices (routers, cross-connects) on the path of the path A procedure for transmitting and receiving control packets between devices, OXC, etc., and securing resources in a communication device and controlling a switch is generally called signaling.

  A communication network such as an optical network constituted by a packet switch network using MPLS, a cross-connect device, OXC, or the like provides data transfer between remote points to the user. When a failure occurs in such a communication network, it is necessary to perform an automatic recovery process and provide a data transfer service without interruption. There is a method of switching data transfer to another path connecting the same two points when data transfer on the path is hindered due to a failure on the communication network.

  In the path failure recovery methods described in Non-Patent Literature 1 and Non-Patent Literature 2, failure recovery is performed by a pre-established failure recovery method and a post-established failure recovery method. The pre-established failure recovery method (1 + 1 Protection, 1: 1 Protection, Shared-Mesh Restoration, etc.) is different from the normal data transfer path (hereinafter referred to as the working path). Alternatively, a plurality of paths (hereinafter referred to as backup paths) are established, and data transfer is switched from the working path to the backup path when a failure occurs on the communication network.

  In addition, the post-establishment type failure recovery method (LSP Restoration or LSP re-routing) does not establish a backup path in advance, but establishes another path (hereinafter referred to as “relief path”) by signaling or the like after the failure occurs, and data transfer Is switched from the working path to the relief path.

CCMP Working Group Internet Draft Expiration Date: November 2003 CCMP GMPLS P & R Design Team by Eric Mannie, Dimitri Papadimitiou May 2003 Recovery (Protection and Restorarion) Terminology for Generalized Multi-Protocol Label Switching (GMPLS) draft-ietf-ccamp-gmpls-recovery- terminology-02.txt CCMP Working Group Internet Draft Expiration Date: April 2005 CCMP GMPLS P & R Design Team JP Lang, Y.Rekhter, D. Papadimitiou October 2004 RSVP-TE Extensions in support of End-to-End Generalized Multi-Protocol Label Switching (GMPLS)- based Recovery draft-ietf-ccamp-gmpls-recovery-e2e-signaling-02.txt

  However, in the above-mentioned prior art failure recovery method, which is the prior art, the backup path is established before a failure occurs on the communication network, so the resources allocated to the backup path are established by other working paths. It is not possible to use it. For this reason, there is a problem in that the backup path becomes a useless state that is not used for data transfer without a normal failure. In addition, there is a problem that the path of the backup path may not always be optimal when the availability of the communication network changes due to the increase / decrease of communication network facilities or the establishment / release of other working / backup paths. . Therefore, the pre-established failure recovery method has a problem that communication efficiency is deteriorated and running cost is increased.

  In addition, in the latter prior art failure recovery method, which is the latter prior art, since the signaling for establishing a rescue path is executed after a failure occurs, it takes time to recover from the failure compared to the pre-established failure recovery method. was there. In addition, when a plurality of working paths including a common link as a part of the route are established, if a failure occurs in the common link, the plurality of working paths establish respective rescue paths by signaling or the like. There must be. At this time, if a rescue path including a link that does not have a sufficient free bandwidth remaining is selected as a plurality of rescue paths corresponding to each active path, the predetermined rescue path is a path due to insufficient free bandwidth of the link. Fails to establish. When such a resource conflict of the rescue path occurs, there is a problem in that it takes time to recover from the failure because it is necessary to try to re-establish the rescue path for the rescue path that has failed to be established due to the conflict. .

  The present invention has been made in view of the above, and an object of the present invention is to obtain a communication device, a communication system, and a communication program that perform reliable and high-speed post-establishment type failure recovery at low cost.

  In order to solve the above-described problems and achieve the object, the present invention provides a communication apparatus in a communication system that performs connection-type communication, wherein the own apparatus transfers data on a working path that is disconnected due to a failure in the communication system. A path storage unit that stores working path information regarding the working path that is a path of the path, and a data transfer path that is set in place of the working path stored in the path storage unit when the working path is disconnected A relief path setting unit that selects and stores relief path information related to the relief path corresponding to the working path before the working path is disconnected, and the relief path setting unit selects when the working path is disconnected. And a failure notification unit that transmits the rescue path information corresponding to the stored working path to the node on the working path that has been disconnected.

  According to the present invention, since the repair path information set in advance is transmitted to the node on the working path that is disconnected after the working path is disconnected, the repair path can be quickly established. Thus, it is possible to perform high-speed post-establishment type failure recovery at low cost.

  Embodiments of a communication apparatus according to the present invention will be described below in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

Embodiment 1 FIG.
FIG. 1 is a diagram showing a configuration of a communication system according to the first exemplary embodiment of the present invention. The communication system 1 is a connection-type communication system that performs post-establishment-type failure recovery, and includes a communication network that connects a plurality of nodes 11 to 22. In the communication system 1, nodes (routers and OXC) 11 to 22 having a data transfer function are connected in a predetermined network pattern. Client nodes 41 and 42 such as computers are connected via nodes 11 to 22. In the first embodiment, the client nodes 41 and 42 transmit and receive data via the nodes 11 to 22. The nodes 11 to 22 here correspond to the communication devices described in the claims.

  For example, the link between the node 11 and the node 12 (link 101 to be described later) is a link for transferring data between the node 11 and the node 12, and physically, for example, by an optical fiber or the like that connects the OXCs. It is configured.

  The links 101 to 103 are links between the nodes 11 and 12, between the nodes 12 and 13, and between the nodes 13 and 14, respectively. The links 104 to 106 are links between the nodes 15 and 16, between the nodes 16 and 17, and between the nodes 17 and 18, respectively. The links 107 to 109 are links between the nodes 19 and 20, between the nodes 20 and 21, and between the nodes 21 and 22, respectively. The links 110 to 113 are links between the nodes 11 and 15, between the nodes 12 and 16, between the nodes 13 and 17, and between the nodes 14 and 18, respectively. The links 114 to 117 are links between the nodes 15 and 19, between the nodes 16 and 20, between the nodes 17 and 21, and between the nodes 18 and 22, respectively.

  In the first embodiment, for convenience of explanation, a description will be made assuming that a node serving as a start point and a node serving as an end point are end point nodes when performing data transfer between the client nodes 41 and 42. A node that performs data transfer when performing data transfer between a start node and an end node will be described as a relay node.

  Here, the path of the communication system 1 will be described. FIG. 2 is a diagram for explaining paths in the communication system. In the first embodiment, working paths (paths for normal data transfer) 140X and 150X are set as paths in the communication system 1, and a failure occurs in a link between nodes 16 and 17 (link 105 described later). A case will be described in which relief paths (another path established by signaling or the like after the occurrence of a failure) 140Y, 140Z, and 150Y are set.

  The working path 140X is a route that reaches the end node 22 from the start node 11 via the relay nodes 12, 16, 17, and 18. The working path 150X is a path that reaches the end node 13 from the start point node 19 via the relay nodes 15, 16, and 17.

  The relief paths 140Y and 140Z are relief paths corresponding to the working path 140X, and the relief path 150Y is a relief path corresponding to the working path 150X. The relief path 140Y is a path that reaches the end point node 22 from the start point node 11 via the relay nodes 12, 13, 14, and 18. The relief path 140Z is a route that reaches the end point node 22 from the start point node 11 through the relay nodes 12, 16, 20, and 21. The relief path 150Y is a path that reaches the end point node 22 from the start point node 19 via the relay nodes 20, 21, and 17.

  Next, the configuration of the nodes 11 to 22 will be described. Since the nodes 11 to 22 have the same configuration, the node 16 will be described as an example here. FIG. 3 is a block diagram showing the configuration of the node. The node 16 includes a control unit 50 that controls control communication for path establishment and failure recovery, route selection processing for path establishment and failure recovery, and transmission / exchange that performs data transfer as an original function of the communication network. Part 60.

  The control unit 50 includes a control communication unit 51, a signaling processing unit 52, a routing processing unit 53, a failure recovery management unit (failure notification unit) 54, a path management unit (path storage unit) 55, a topology database 56, and a rescue path route selection unit. A (relief path setting unit) 57 and a resource management unit 58 are provided.

  The control communication unit 51 has a function of transmitting / receiving a control packet for signaling to / from another node, and is connected to a control communication network that transfers the control packet. Nodes other than the node 16 (nodes 11 to 15 and 17 to 22) are also connected to the control communication network. The control communication network is a communication network that transfers control packets for signaling and routing and a failure recovery notification packet to be described later, and may be realized by using the communication network shown in FIG. 1 or by using another communication network. It may be realized.

  The signaling processing unit 52 processes a signaling procedure for path establishment and release, and transmits and receives signaling control packets via the control communication unit 51. In addition, the signaling processing unit 52 transmits information on the working path acquired by the signaling procedure to the path management unit 55. The routing processing unit 53 processes a routing protocol for collecting the network topology, and updates information in the topology database 56.

  The failure recovery management unit 54 manages information regarding a working path that is disconnected when a failure occurs and information on a rescue path corresponding to each working path for each possible failure. The failure recovery management unit 54 notifies the path information of the rescue path to, for example, the start point nodes 11 and 22 when a failure occurs. When the start node (here, the nodes 11 and 19) receives information including the path information of the rescue path from the other nodes, the path management unit 55 and the signaling processing unit 52 are based on the path information included in the received information. Thus, it has a function of establishing a relief path and releasing a working path that has become an obstacle.

  The path management unit 55 manages information related to the working path and the rescue path that are being established or have been established. The path management unit 55 stores a path management table (working path information) 501 described later. The topology database 56 manages information such as adjacency relationships between nodes on the communication network, interface types of links between nodes, and free bandwidth information collected by the routing processing unit 53. The relief path route selection unit 57 has a function of selecting a relief path route using the topology database 56.

  The resource management unit 58 manages the state of resources in the transmission / exchange unit 60 necessary for providing a data transfer function related to the established working path or rescue path. Resource statuses managed by the resource management unit 58 include failure status, operation status, bandwidth used by the working path or rescue path, unused bandwidth, switch setting status, etc. It is.

  The transmission / exchange unit 60 performs data transfer between adjacent nodes. The transmission / exchange unit 60 includes a monitoring control unit 61, demultiplexing units 62A and 62B, multiplexing units 63A and 63B, network interface units 64A to 64D, client interface units 65A and 65B, and a switch 66.

  The demultiplexing units 62A and 62B receive light that is wavelength-multiplexed and transferred in an optical fiber that is a path between nodes, and divides the received light into optical signals of individual wavelengths. The demultiplexing units 62A and 62B transfer the divided optical signals to the network interface units 64A and 64B and the network interface units 64C and 64D, respectively.

  The multiplexing unit 63A wavelength-multiplexes the optical signals of individual wavelengths transmitted from the network interface units 64A and 64B and transfers them onto the path (optical fiber). The multiplexing unit 63B wavelength-multiplexes the individual wavelength optical signals transmitted from the network interface units 64C and 64D and transfers them onto the path.

  The network interface units 64A and 64B and the network interface units 64C and 64D convert the optical signals received from the demultiplexing unit 62A and the demultiplexing unit 62B, respectively, into electric signals and input them to the switch 66. The network interface units 64A and 64B and the network interface units 64C and 64D convert the electrical signals transmitted from the switch 66 into optical signals and transfer them to the multiplexing unit 63A and the multiplexing unit 63B, respectively.

  The client interface units 65A and 65B are interfaces that communicate with client nodes. The switch 66 transfers the electrical signal received from the predetermined network interface units 64A to 64D to the other network interface units 64A to 64D based on the information set in the monitoring control unit 61.

  The monitoring control unit 61 directly monitors and controls the status of the network interface units 64A to 64D, the switch 66, and the like. In FIG. 3, the connection between the monitoring control unit 61 and the network interface units 64B and 64D is omitted. With these functions, for example, an optical signal received by the node 16 from the node 12 through the working path 140X is transferred to the node 17.

  Next, a processing procedure of the communication system 1 will be described. First, the outline of the processing procedure of the communication system 1 will be described with reference to FIG. 4, and then each process will be described in detail. FIG. 4 is a sequence diagram illustrating an outline of a processing procedure of the communication system according to the first embodiment. First, according to the routing protocol, all the nodes 11 to 22 on the communication network transmit the link information of the own node and the link information received from the other adjacent nodes to the adjacent nodes, and the link information from the adjacent nodes. Receive. Each of the nodes 11 to 22 collects topology data of the communication network by the function of this routing protocol (step S401).

Next, the working path 150X is established by signaling (step S402).
In this signaling, a signaling control packet is transmitted from the node 19 to the node 13 via the nodes 15, 16, and 17. In addition, a signaling control packet is transmitted from the node 13 to the node 19 via the nodes 17, 16, and 15 through a path opposite to the control packet path. Resources are secured and switch control is performed by transmission and reception of control packets by each node, and a communication path is established.

  Specifically, signaling includes transmission of a control packet from the node 19 to the node 15, transmission of a control packet from the node 15 to the node 16, transmission of a control packet from the node 16 to the node 17, and transmission from the node 17 to the node 13. Control packet transmission, node 13 to node 17 control packet transmission, node 17 to node 16 control packet transmission, node 16 to node 15 control packet transmission, node 15 to node 19 control This is done by sending a packet.

  After the working path 150X is established, each node 19, 15, 16, 17, 13 on the working path 150X selects a rescue path corresponding to the working path 150X, and stores information on the selected rescue path. Here, since a case where a failure occurs in the link 105 will be described, a case where the node 16 selects and stores the relief path 150Y will be described (step S403).

Next, the working path 140X is established by signaling (step S404).
Signaling here is performed by transmitting control packets to the nodes 11, 12, 16, 17, 18, and 22.

  After the working path 140X is established, each of the nodes 11, 12, 16, 17, 18, and 22 on the working path 140X selects a rescue path corresponding to the working path 140X and stores information on the selected rescue path. Here, since a case where a failure occurs in the link 105 will be described, a case where the node 16 selects and stores the relief paths 150Y and 140Y will be described (step S405).

  Thereby, data transfer is performed by the working paths 140X and 150X. When a failure occurs in the link 105 at a predetermined timing, the node 16 detects the failure (step S406).

  When the node 16 detects the failure of the link 105, the node 16 transmits the stored route information of the rescue path 140Y to the start node 11 of the working path 140X. In addition, the node 16 transmits the stored path information (failure recovery notification) of the rescue path 150Y to the start node 19 of the working path 150X (step S407). Here, when the node 16 detects the failure of the link 105, the stored path information of the rescue paths 140Y and 150Y is respectively stored in the starting node 11 of the working path 140X and the starting node 19 of the working path 150X. Although the case of transmitting will be described, the node 16 may transmit the route information of the rescue paths 140Y and 150Y to any nodes on the working paths 140X and 150Y that are the starting points of establishment of the respective rescue paths.

  When the start node 11 receives the route information from the node 16, the start node 11 establishes a rescue path 140Y corresponding to the route information by signaling (step S408). Further, when receiving the route information from the node 16, the start node 19 establishes a rescue path 150Y corresponding to this route information by signaling (step S409). In addition, the process of step S408 and the process of step S409 are implemented simultaneously in parallel.

  Thereafter, a process of transferring data from the node 11 to the node 22 through the newly established relief path 140Y is performed. In addition, a process of transferring data from the node 19 to the node 13 through the newly established relief path 150Y is performed.

  Next, the processing procedure of the communication system shown in FIG. 4 will be described in detail with reference to FIGS. First, information in the topology database 56 of the node 16 and a method for collecting and updating this information will be described.

(Collecting topology data)
The topology database 56 stores characteristic information such as identification information of each link on the communication network, connection destination node, free bandwidth information, cost of each link, and interface type. These pieces of information are collected by a routing protocol. That is, each of the nodes 11 to 22 on the communication network transmits information related to the link accommodated in the own node to the adjacent node and forwards information related to the link of the other node received from the adjacent node to another adjacent node. Thus, the information of the links accommodated by the nodes 11 to 22 is distributed to all the nodes on the communication network.

  In addition, each node 11 to 22 always transmits the link information of its own node to the adjacent node, and constantly updates the topology database 56 in its own node based on the link information of the other node received from the adjacent node. The latest topology information in the network is maintained at each node. The collection of topology data by this routing protocol corresponds to the processing of step S401 described in the sequence diagram of FIG.

  By using the topology database 56, each of the nodes 11 to 22 can determine whether a path can be established between desired nodes and which path can be established.

  As a routing protocol here, for example, OSPF-TE (Open Shortest) shown in Internet Engineering Task Force (IETF) Internet draft (draft-ietf-ccamp-ospf-gmpls-extensions-12.txt, RFC3630) and the like. Path First-Traffic Engineering).

(Establish working path)
Next, a process for establishing working paths 140X and 150X by signaling for data transfer will be described. Each of the nodes 11 to 22 transfers a signaling control packet A on the control communication network in order to establish a path.

  The control packet A includes start node identification information (start node identification information), end node identification information (end node identification information), and path identification information (path identification information). Further, the control packet A includes constraints for guaranteeing end-to-end communication quality such as interface type, required bandwidth and priority. Further, the control packet A includes information regarding whether or not recovery is necessary when a failure occurs. Further, the control packet A may include route information from the start node to the end node.

  The path information includes node identification information on the path, link identification information between nodes on the path, resource identification information on links (MPLS and GMPLS labels), and the like. The resource on the link here is one wavelength multiplexed on the fiber in the WDM network, one time slot in the SONET / SDH network, and the value of the label attached to the packet header in the packet switch network. is there.

  When the node adjacent to the start point node receives the control packet A from the start point node, it determines whether or not the node can provide the requested path based on whether or not the resource is available. Then, the control packet A is transferred to another adjacent node different from the node that has transmitted the control packet.

  When each relay node includes route information for determining a transfer destination adjacent node in the control packet A, the relay node determines a transfer destination adjacent node based on the route information and includes the route information. If not, an appropriate adjacent node is selected using the topology database 56.

  When the control packet A reaches the end point node by repeating the transfer process of the control packet A, the end point node transmits the control packet B toward the reverse route to which the control packet A has been transferred. This control packet B includes information such as start node identification information, end node identification information, path identification information, a link to be used, and a resource (label).

  Each relay node that receives this control packet B makes preparations necessary for data transfer, such as allocation of resources on the link (selection of wavelength to be used and determination of label value), switch setting, etc., and transfer to adjacent nodes To do. The control packet B reaches the start point node by repeating the transfer process of the control packet B.

  As a result, the working path is established, and the start node starts data transfer on the working path toward the end node. The transfer procedure of these control packets A and B corresponds to the processing of steps S402 and S404 described in the sequence diagram of FIG. That is, the working paths 140X and 150X are established by the transfer processing of the control packets A and B. Then, the start node 11 starts data transfer to the end node 22, and the start node 19 starts data transfer to the end node 13.

  In addition, when transferring control packet A or control packet B, each node (start node, relay node, end node) on the path has start node identification information, end node identification information, and constraint conditions (necessary bandwidth, interface type, etc.) The path information and the like are acquired (extracted) from the information in the control packet and stored in the path management unit 55.

  FIG. 5 is a diagram illustrating an example of a configuration of a path management table stored in the path management unit. As shown in the figure, the path management table 501 includes path identification information for each path (path # 1, path # 2, etc.), start node identification information, end node identification information, interface type, necessary bandwidth, priority, and failure recovery. Need, route information, and the like.

  Here, the path identification information of path # 1 is the identification information of working path 150X, the start node identification information is identification information of node 19, the end node identification information is identification information of node 13, and the interface type is “2 (Ethernet (registered trademark)”. )) ”, The case where the required bandwidth is 10 Gbps, the priority is“ 1 ”, and the necessity of failure recovery is“ 1 (requires recovery) ”. Further, as the route information, the case where the order of the data transfer route is node 19, link 114, node 15, link 104, node 16, link 105, node 17, link 112, and node 13 (working path 150X) is shown. Yes.

  As a signaling protocol of the working paths 140X and 150X, for example, RSVP-TE (Resource reSerVation Protocol-Traffic Engineering) shown in RFC3473 and 3209 of IETF is used. In RSVP-TE, the control packet A corresponds to a “Path message”, and the control packet B corresponds to a “Resv message”.

  Acquisition of path information of a path established at the time of signaling is realized as a function of Route Recording using RRO (Record Route Object) in the RSVP-TE signaling protocol, and signaling control packets A and B at the time of path establishment Node identification information on the route, link identification information between nodes on the route, and resource identification information on the link are sequentially added and transferred.

(Relief path selection)
Next, a description will be given of a process in which each node on the working path route selects a rescue path route corresponding to the working path when the working path is established. This relief path selection processing corresponds to the processing of steps S403 and S405 described in the sequence diagram of FIG.

  FIG. 6 is a flowchart illustrating the processing procedure of the relief path selection process. The failure recovery management unit 54 of the node 16 for each possible failure (failure for each link), list information of the working path that is disconnected when the failure occurs, and information of the rescue path corresponding to each working path Is managing.

  The failure recovery management unit 54 of the node 16 stores the working paths 140X and 150X that are disconnected due to the failure of the link 105 and the working path 150X that is disconnected due to the failure of the link 104 in the list information.

  Here, a case where a single link failure occurs will be described as an example of the failure. A link failure is a case where a transmission path between two adjacent nodes is disconnected, and an example of this is disconnection of an optical fiber in a WDM network. For example, when the optical fiber is cut, the optical signal transmitted on the optical fiber does not reach the adjacent node. As a result, all the wavelength paths transferred on the optical fiber become obstacles, and normal data transfer is prevented. That is, the paths that need to be relieved when such a link failure occurs are all the working paths established using the failed link.

  A single link failure here refers to a situation where one link has failed and no other failure has occurred. Therefore, the only path that needs to be relieved when a single link failure occurs is the working path that uses the failed link, and other working paths that do not have a failure need to be relieved. Absent.

  For example, when the working path 150X is established (step S801), the failure recovery management unit 54 of the node 16 sets the working path 150X using the link 105 based on the information held by the path management unit 55. Recognize that As a result, the failure recovery management unit 54 determines that the working path 150X needs to be relieved when the link 105 fails (step S802).

  Then, the failure recovery management unit 54 extracts, from the information stored in the path management unit 55 (path management table 501), information on all the working paths that use the link 105 as the failure recovery content at the time of failure of the link 105. . The working path extracted by the failure recovery management unit 54 includes a previously established path and a newly established working path.

  The rescue path route selection unit 57 selects the rescue path at the time of failure of the link 105 only for each of the active paths extracted by the failure recovery management unit 54 and requiring failure recovery. That is, the rescue path route selection unit 57 selects a list of working paths that are affected by the failure (step S803). Here, the rescue path route selection unit 57 uses the topology database 56 collected and stored by the routing protocol, the information on the working path stored by the path management unit 55, and the information on the link 105 to use the rescue path. Selection is started (steps S804 and S805).

  Then, information on all links used by the working path is added to the topology database 56 (step S806). In other words, in addition to the availability on the current network (before the failure), it is assumed that the bandwidth of the link currently used by the working path is also free (that can be used for the rescue path after the failure) Add to topology database 56. This is because the link used by the working path is not used as the working path when data transfer is switched to the relief path, and can be used for the relief path. The working path here may include not only the single working path selected in the process of step S805 but also other working paths that are disconnected due to a failure in the communication system 1. As a result, the number of link candidates to be used as the path of the rescue path is increased, the possibility that a rescue path can be established is increased, and the number of newly used links is decreased.

  Furthermore, if the cost of the link used by the working path is set to a small value when selecting a route for one rescue path, the possibility of using this low cost link for the rescue path increases. Therefore, the number of common paths increases between the working path and the relief path. In a node that takes time to establish a link and establish a cross-connect state (setting of the switch 66) at the time of establishing a rescue path, the time is shorter than when the cost of the link used by the working path is not set to a small value. This increases the possibility that a relief path can be established. In other words, when a link that is completely different from the working path is used, it takes time to reserve a link and establish a cross-connect. On the other hand, by using the link used by the working path as a rescue path, the link reservation and crossing are performed. Connection can be established in a short time, which saves time.

  Next, since it becomes impossible to relieve if a link having a failure in the rescue path is used, information on the link 105 that becomes a failure is deleted from the topology database 56 (step S807).

  Next, the rescue path route selection unit 57 selects the route of the rescue path at the time of link failure based on the new topology database 56 obtained by the processes of steps S806 and S807 (step S808). As a method for selecting a relief path, a method called a Dijkstra method is generally known. Here, as a method for selecting a relief path, for example, the selection is made by the Dijkstra method.

  In this Dijkstra method, it is possible to obtain a route that minimizes the total cost of the links on the route from the start point node to the end point node based on the topology information. By using this Dijkstra method, it is possible to obtain a route that satisfies the constraints (required bandwidth, interface type, etc.) of the working path. The rescue path route selection unit 57 determines whether there is a rescue path route at the time of a link failure (step S809). When the rescue path route selection unit 57 determines that there is a rescue path route at the time of link failure (Yes in step S809) and selects a predetermined rescue path, the rescue path management unit 54 selects the rescue path selected. The route is stored (step S811).

  If there is not enough free bandwidth, there is a possibility that no relief path exists. In this case, a relief path cannot be established. If the rescue path route selection unit 57 determines that there is no rescue path route at the time of the link failure (No in step S809), the node 16 indicates that the rescue path cannot be established at the current node as the starting point node. A notification packet is transmitted (step S810). As a result, the start node can recognize in advance that the working path is not relieved when a failure occurs.

  After selecting a rescue path for one working path, the bandwidth of the link used by the rescue path selected immediately before is removed from the topology database 56 (step S812). Then, after selecting a rescue path for one working path, the process of selecting a rescue path for the next working path is repeated (steps S805 to S813). At this time, similar to the processing in step S808, the path of the rescue path is selected by, for example, the Dijkstra method. By repeating these processes, it becomes possible to select the path of the rescue path for all the working paths.

  The working path 150X also uses the link 104. The node 16 also selects a rescue path for the link 104 by the same processing as that for the link 105 (step S814).

  Also, when the topology database 56 is changed by the routing protocol or when the working path is released (step S801), the rescue path route selection unit 57 again executes the process of selecting a rescue path (steps S802 to S802). S814). As a result, the failure recovery management unit 54 can always store optimal rescue path information.

  As described above, the failure recovery management unit 54 selects the rescue path route of the working path that is affected when a failure occurs for all the links accommodated by the node, and stores the selection result. The failure recovery management unit 54 stores information regarding the route of the selected rescue path as the failure recovery management table 502.

  FIG. 7 is a diagram illustrating an example of the configuration of the failure recovery management table. As shown in the figure, the failure recovery management table 502 includes a failure type for each failure (failure # 1, failure # 2, etc.), failure identification information, the number of active paths affected by the failure, and one to a plurality of failure recovery information. (Failure recovery information # 1 to #n (n is a natural number)). Fault recovery information is included in the same number as the number of working paths affected by the fault. For example, when there are two working paths 140X and 150X that are affected by the failure of the link 105, the failure recovery information is composed of two. Each failure recovery information includes working path identification information and rescue path route information indicating the route of the rescue path.

  In this example, the failure type of failure # 1 is “1 (link failure)”, the failure identification information is link 105, and the number of working paths affected by the failure is two. Further, as an example of the failure recovery information, a case is shown in which the working path identification information of the failure recovery information # 1 is the working path 150X, and the rescue path route information is the route of the rescue path 150Y. The relief path route information (route of the relief path 150Y) here indicates that the order of the data transfer route is node 19, link 107, node 20, link 108, node 21, link 116, node 17, link 112, and node 13. It shows a case.

  Although the case where the failure is a single link failure has been described in the first embodiment, the failure target managed by the node 16 is not limited to a single link failure. For example, when a plurality of optical fibers or the like are physically housed in the same conduit and embedded in the ground, there is a high possibility that the plurality of optical fibers will be cut simultaneously due to construction or disaster. In such a case, a single failure factor may cause a plurality of links to be disconnected, and the node 16 may select the path of the rescue path for all the working paths included in these links.

  In addition, when an adjacent node becomes a failure due to a power failure or the like, a rescue path route may be selected for all working paths accommodated by the adjacent node. In addition, when the predetermined link accommodated by the predetermined adjacent node and the own node is grouped in advance, when any adjacent node or link in the group becomes a failure, the adjacent node and link in the group are routed. The path of the rescue path may be selected with respect to the working path.

  As described above, by managing each working path and the rescue path corresponding to this working path for each possible failure factor, it becomes possible to perform optimum failure recovery when each failure occurs. .

  Such establishment processing of the working path, selection processing of the rescue path, and storage of information regarding the selected rescue path are performed when the working paths 140X and 150X are established. First, when signaling for establishing the working path 150X is performed, the path selection of the rescue path when the link 105 becomes faulty is performed in the node 16, and the path of the rescue path 150Y of the working path 150X is determined. The path of the rescue path 150Y is a path from the start point node 19 to the end point node 13 via the relay nodes 20, 21, and 17, as shown in FIG.

  Note that there is not always only one rescue path route candidate. In other words, depending on the route calculation algorithm and the cost of the link between nodes, another route of the rescue path may be selected. For example, the rescue path 150Y passes through three relay nodes, but the route from the start node 19 to the end node 13 through the relay nodes 20, 16, 12 also includes three relay nodes as in the rescue path 150Y. Will go through. For this reason, if there is a sufficient free bandwidth in the link 111 between the node 16 and the node 12, this route can be used as the route of the rescue path.

  Next, when signaling for establishing the working path 140X is performed, the path selection of the rescue path when the link 105 becomes faulty is performed in the node 16. The working paths affected when the link 105 fails are 140X and 150X.

  When the path of the rescue path 150Y of the working path 150X is selected first, the next time the path of the rescue path of the working path 140X is selected, the information on the link used by the rescue path 150Y is removed from the topology data. Since a path is selected, there is no conflict of relief paths. That is, when the rescue path 150Y is selected as the rescue path of the working path 150X and the rescue path 140Z is selected as the rescue path of the working path 140X, a conflict occurs between the rescue path 140Z and the rescue path 150Y in the link 108. By selecting the relief path 150Y, no conflict of relief paths occurs in the communication system 1.

  For example, if there is no free bandwidth in the link 108 or the like, a route such as the rescue path 140Z (from the start node 11 to the end node 22 via the relay nodes 12, 16, 20, and 21) is not selected. As a result, the route of the rescue path 140Y is selected. On the other hand, if there is no free bandwidth in the link 117 or the like, the route such as the rescue path 140Y is not selected, and the route of the rescue path 140Z is selected.

(Failure notification processing)
Next, failure notification processing when a failure occurs will be described. The failure detection processing and failure notification processing here correspond to the processing of steps S406 and S407 described with reference to the sequence diagram of FIG. The node 16 that has detected the failure notifies the start point node (start point nodes 11 and 19) of the working path of a notification packet for starting the failure recovery. The node 16 detects a link failure by a predetermined means. For example, in a communication network that switches light transparently, the node 16 uses a method of using the LMP (Link Management Protocol) Fault Management function (IETF Internet draft (draft-ietf-ccamp-lmp-09.txt)). It is possible to detect a link failure.

  FIG. 8 is a flowchart showing a processing procedure of failure recovery notification processing when a single link failure is detected. When the node 16 detects a single link failure (link 105) (step S901), the failure recovery management unit 54 extracts a list of working paths and rescue path routes using the link 105 from the failure recovery management table 502 (step S901). S902).

  For each of the extracted active paths (in this case, the active paths 140X and 150X), the failure recovery management unit 54 starts the path from the path management table 501 and the failure recovery management table 502 in the path management unit 55 to the start node and the relief path. Information is extracted to create a notification packet (steps S903 to S906).

  For example, the failure recovery management unit 54 extracts the route information of the start point node 19 and the rescue path 150Y and creates a notification packet (failure recovery notification). The created notification packet is transmitted toward the start node 19 (step S907). The notification packet includes identification information of the own node, failure type (link failure), detailed failure information (failure factors such as optical input interruption, failure link identification information, etc.), failure detection time, working path identification information, relief path Include the route etc. The failure recovery management unit 54 repeats the process of creating a notification packet for each working path extracted from the failure recovery management table 502 (steps S902 to S908).

  Here, the configuration of the failure recovery notification packet (failure recovery notification message) will be described. FIG. 9 is a diagram illustrating an example of the configuration of the failure recovery notification packet. As shown in the figure, a failure recovery notification packet 503 includes a packet (message) type, a notification number, a transmission source node, a failure type, detailed failure information, failure identification information, a failure occurrence time (failure detection time), and recovery. It is configured including the number of paths and 1 to a plurality of failure recovery information (failure recovery information # 1 to #n).

  Here, the packet type is “1 (failure recovery notification)”, the notification number is “1001”, the transmission source node is node 16, the failure type is “1 (link failure)”, and the detailed failure information is “3 (optical input). ”)”, The failure identification information is the link 105, the failure occurrence time is February 1, 2005, 15:00:00, and the number of recovery paths is “1”. The failure recovery information # 1 here is the same information as the failure recovery information # 1 described in FIG.

  Note that the node 16 may transmit a failure recovery notification packet 503 for each active path, or create a single failure recovery notification packet 503 by grouping a plurality of active paths having the same start node. It may be sent to the node.

  Next, the relief path establishment process of the start point node when the failure recovery notification packet 503 is received will be described. The relief path establishment processing here corresponds to the processing in steps S408 and S409 described in the sequence diagram of FIG.

(Establishing a relief path)
FIG. 10 is a flowchart showing the processing procedure of the relief path establishment process. Here, since the start node 11 and the start node 19 that have received the failure recovery notification packet 503 from the node 16 operate according to the same processing procedure, the processing procedure of the starting node 19 will be described as an example.

  When receiving the failure recovery notification packet 503 transmitted from the node 16 (step S1001), the starting point node 19 of the working path starts a process of establishing a rescue path based on information in the notification packet. The start node 19 extracts information on the working path (working path identification information) and information on the path of the rescue path (rescue path path information) included in the failure recovery notification packet 503 (steps S1002 and S1003).

  Based on the extracted working path identification information and rescue path route information, the start point node 19 adds information about the route to the signaling control packet and transmits it to the adjacent node on the route (step S1004). For example, in RSVP-TE, the path for transmitting the signaling control packet is transferred in the ERO (Explicit Route Object) format.

  When each relay node (nodes 20, 21, and 17) receives the control packet, it reads out information related to the path included in the control packet, determines the node (adjacent node) to which the next control packet is transferred, and transfers it. As each relay node repeats this process, the control packets are sequentially transferred to the nodes on the route and reach the end node 13.

  Thereafter, the order of the start point node 19, the relay node 20, the relay node 21, the relay node 17, and the end point node 13 is reversed (the order of the end point node 13, the relay node 17, the relay node 21, the relay node 20, and the start point node 19). The control packet is transferred through the route, and the control packet reaches the start node 19. As a result, a relief path 150Y is established (step S1005). The start of the relief path establishment by the start point node is performed for all the relief paths included in the failure recovery notification packet 503 (steps S1002 to S1005).

  When the relief path corresponding to the start point node 19 is established (step S1006), the start point node 19 switches the data transfer on the relief path 150Y (step S1007). Then, the working path 150X where the failure has occurred is released. Based on the path management table 501 in the path management unit 55 (step S1008), the start node 19 sends a control packet for releasing the working path 150X to an adjacent node (relay nodes 20, 21, 17) on the working path 150X. (Step S1009). The control packets are sequentially transferred to the nodes on the path of the working path 150X, and the working path is released by reaching the end node 13. That is, each node that receives the control packet from the start node 19 releases the working path 150X by the signaling processing unit 52 and the resource management unit 58.

  When the relief path establishment process and the relief path release process are completed in each node, the path management unit 55 of each node updates and stores the path management table 501 in the latest state (step S1010). The relief path is established by the start node for all the relief paths included in the failure recovery notification packet 503 (steps S1006 to S1011).

  In the first embodiment, the path information of the rescue paths 140Y and 150Y may include information on nodes, links, and resources on the link that are the path of the rescue path, and should be passed as the path of the rescue path. Information about non-nodes, links, and resources on links may be included.

  Here, the starting point node 19 establishes the relief path 150Y and then releases the working path 150X. However, after the starting point node 19 releases the working path 150X, the relief path 150Y may be established.

  In the first embodiment, the failure recovery management unit 54, the rescue path route selection unit 57, and the topology database 56 are included in one node. However, each of these may be a physically different computer or the like. A configuration may be adopted in which the failure is notified to the failure recovery management unit 54 from a node in which another failure is detected or another device. In the first embodiment, the node 16 performs processing such as path selection of a rescue path for the failure of the link 105, but the node 17 performs processing such as path selection of a rescue path. Also good.

  As described above, according to the first embodiment, since the backup path is not established before the failure occurs, management of the backup path becomes unnecessary, and the operation of the communication network is facilitated. In addition, establishment of another working path is not hindered by the backup path. In addition, after establishing a backup path, the topology of the communication network changes due to the addition or removal of a communication network, the establishment / release of another working path, etc., and the rescue path is not optimal (for example, it is not the shortest path). There is no.

  In addition, since the nodes that have detected the failure are collectively selected for the relief path so that the relief path is not concentrated on a specific resource (conflict occurs), no conflict occurs. Therefore, reliable and high-speed failure recovery is possible. In addition, since the rescue path is selected before the occurrence of the failure, the failure can be recovered at high speed.

  In addition, since the path selection processing of the rescue path and the failure recovery notification transmission are performed only for the fault related to the node adjacent to the fault location (link), the amount of data is small and the path selection is small regardless of the scale of the communication network. Fault recovery can be easily performed with a large amount of calculation. Furthermore, since a node adjacent to the failure location detects a failure and transmits a failure recovery notification, time for notifying the management device or the like of the failure detection is unnecessary, and high-speed failure recovery is possible.

  Further, by applying the communication system 1 to the GMPLS or MPLS signaling protocol and routing protocol, it becomes possible to perform failure recovery only by extension for notifying the route of failure recovery, and therefore, simple and low-cost extension It becomes possible to perform failure recovery with. In particular, in GMPLS, since the procedure for failure notification is standardized, it is possible to perform failure recovery only by adding the path information of the rescue path to this.

Embodiment 2. FIG.
Next, a communication system according to the second embodiment will be described with reference to FIG. In the second embodiment, the node selects a rescue path after detecting a failure. That is, in the first embodiment, every time the working path is established / released or the topology database 56 is updated, the rescue path selection and storage processing described in FIG. 6 is executed. When establishing / releasing the working path or updating the topology database 56, the process of selecting and saving the rescue path is not executed. In the second embodiment, after detecting a failure, only a link that has failed is subjected to processing for selecting and saving a rescue path and transmitting a failure recovery notification.

  FIG. 11 is a sequence diagram of the processing procedure of the communication system according to the second embodiment. In addition, the description about the procedure which performs the process similar to Embodiment 1 is abbreviate | omitted, and it demonstrates centering around the characteristic process procedure of the communication system 1 concerning Embodiment 2. FIG.

  First, each of the nodes 11 to 22 collects communication network topology data using a routing protocol (step S1101). Next, the working path 150X is established by signaling (step S1102). Next, the working path 140X is established by signaling (step S1103).

  Thereby, data transfer is performed by the working paths 140X and 150X. When a failure occurs in the link 105 at a predetermined timing, the node 16 detects the failure (step S1104).

  The node 16 on the working path 150X selects the rescue path 150Y corresponding to the working path 150X. Further, the node 16 on the working path 140X selects the rescue path 140Y corresponding to the working path 140X (step S1105).

  The node 16 transmits the route information of the selected rescue path 150Y to the start node 19 of the working path 150X as a failure recovery notification. Further, the node 16 transmits the route information of the selected rescue path 140Y as a failure recovery notification to the start node 11 of the working path 140X (step S1106).

  Upon receiving the route information from the node 16, the start node 11 establishes a rescue path 140Y corresponding to this route information by signaling (step S1107). Further, when receiving the route information from the node 16, the start node 19 establishes a rescue path 150Y corresponding to the route information by signaling (step S1108).

  As a result, the communication system 1 according to the second embodiment does not need to always hold information on the rescue path according to the cause of the failure, as compared with the communication system 1 according to the first embodiment. The processing load for path selection is reduced, and the mounted memory of each node can be reduced.

  As described above, according to the second embodiment, even when the topology state of the communication network frequently changes and there are many processes for selecting a rescue path, reliable and fast post-establishment type fault recovery is possible. Can be performed at low cost.

  As described above, the communication device, the communication system, and the communication program according to the present invention are suitable for the communication network failure recovery.

It is a figure which shows the structure of the communication system concerning Embodiment 1 of this invention. It is a figure for demonstrating the path | route in a communication system. It is a block diagram which shows the structure of a node. FIG. 3 is a sequence diagram illustrating an outline of a processing procedure of the communication system according to the first embodiment. It is a figure which shows an example of a structure of the path management table which a path management part memorize | stores. It is a flowchart which shows the process sequence of the selection process of a relief path. It is a figure which shows an example of a structure of a failure recovery management table. It is a flowchart which shows the process sequence of the failure recovery notification process at the time of detecting a single link failure. It is a figure which shows an example of a structure of a failure recovery notification packet. It is a flowchart which shows the process sequence of a relief path establishment process. FIG. 10 is a sequence diagram illustrating a processing procedure of the communication system according to the second exemplary embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Communication system 11-22 Node 41, 42 Client node 50 Control part 51 Control communication part 52 Signaling process part 53 Routing process part 54 Fault recovery management part 55 Path management part 56 Topology database 57 Relief path route selection part 58 Resource management part 60 Transmission / exchange unit 61 Monitoring control unit 62A, 62B Demultiplexing unit 63A, 63B Multiplexing unit 64A-64D Network interface unit 65A, 65B Client interface unit 66 Switch 101-117 Link 140X, 150X Working path 140Y, 140Z, 150Y Rescue path 501 Path management table 502 Failure recovery management table 503 Failure recovery notification packet

Claims (14)

In a communication device in a communication system that performs connection-type communication,
A path storage unit that stores working path information regarding a working path in which the own apparatus is a data transfer path in a working path that is disconnected due to a failure in the communication system;
When the working path is disconnected, the working path connects the saving path information related to the working path corresponding to the working path set as the data transfer path instead of the working path stored in the path storage unit. A relief path setting unit that selects and stores the information before it is interrupted,
When the working path is disconnected, a failure notifying unit that transmits relief path information corresponding to the working path selected and stored by the rescue path setting unit to a node on the disconnected working path;
A communication apparatus comprising: In a communication device in a communication system that performs connection-type communication,
A path storage unit that stores working path information regarding a working path in which the own apparatus is a data transfer path in a working path that is disconnected due to a failure in the communication system;
When the working path is disconnected, the rescue path setting for selecting and storing relief path information related to the rescue path corresponding to the working path set as a data transfer path instead of the working path stored in the path storage unit And
A failure notification unit that transmits the rescue path information corresponding to the working path selected and stored by the rescue path setting unit to a node on the working path that has been disconnected;
A communication apparatus comprising:   The communication apparatus according to claim 1, wherein the working path information in the path storage unit is information related to a working path that is disconnected due to a single link failure in the communication system.   The communication apparatus according to claim 1, wherein the working path information in the path storage unit is information related to a working path that is disconnected due to a single node failure in the communication system.   2. The working path information in the path storage unit is information on a working path that is disconnected due to a grouped failure including a predetermined link failure and a predetermined node failure in the communication system. Or the communication apparatus of 2.   The communication apparatus according to claim 1, wherein the rescue path information includes information regarding a node, a link, and / or a resource on the link that is a path of the rescue path.   The communication apparatus according to claim 1, wherein the rescue path information includes information regarding a node, a link and / or a resource on the link that should not pass as a path of the rescue path.   When the rescue path setting unit cannot select the rescue path information, the failure notification unit displays information indicating that the rescue path information could not be selected before the working path is disconnected. The communication apparatus according to claim 1, wherein the rescue path information is transmitted to a node on the working path where the rescue path information could not be selected.   9. The rescue path setting unit, when selecting and storing the rescue path information, selects and stores the rescue path information using a resource that is being used by an active path. The communication apparatus as described in any one.   The rescue path setting unit selects and stores the rescue path information by using the resource being used by the working path preferentially over other resources when selecting and storing the rescue path information. The communication apparatus according to any one of claims 1 to 8.   When the rescue path setting unit selects and stores the rescue path information, another working path that is disconnected due to a failure in the communication system that is different from the working path that is the target of the rescue path information is being used. 9. The communication apparatus according to claim 1, wherein the rescue path information is selected and stored using a resource.   The communication apparatus according to claim 1, wherein the failure notification unit transmits a plurality of the relief path information to a node on the working path that is disconnected. . In a communication system in which a plurality of communication devices perform connection-type communication,
The communication device
When the working path that is the data transfer path of the own device is disconnected due to a communication failure, the working path is connected to the rescue path information related to the rescue path that is set as the data transfer path instead of the current path. When the working path is disconnected, the rescue path information corresponding to the selected and stored working path is transmitted to the communication device on the disconnected working path.
The communication device on the working path that has been disconnected establishes the rescue path based on the rescue path information. A process of storing working path information relating to a working path in which the own apparatus is a data transfer path in a working path that is disconnected due to a failure in a communication system that performs connection-type communication;
When the working path is disconnected, the working path connects the saving path information related to the working path corresponding to the working path set as the data transfer path instead of the working path stored in the path storage unit. Processing to select and store before declining,
When the working path is disconnected, the process of transmitting the repair path information corresponding to the working path selected and stored by the rescue path setting unit to a node on the disconnected working path is sent to the computer. A communication program to be executed.

Images